Evaluation of the Intracellular Signaling Activities of κ-Opioid Receptor Agonists, Nalfurafine Analogs; Focusing on the Selectivity of G-Protein- and β-Arrestin-Mediated Pathways

Opioid receptors (ORs) are classified into three types (μ, δ, and κ), and opioid analgesics are mainly mediated by μOR activation; however, their use is sometimes restricted by unfavorable effects. The selective κOR agonist nalfurafine was initially developed as an analgesic, but its indication was changed because of the narrow safety margin. The activation of ORs mainly induces two intracellular signaling pathways: a G-protein-mediated pathway and a β-arrestin-mediated pathway. Recently, the expectations for κOR analgesics that selectively activate these pathways have increased; however, the structural properties required for the selectivity of nalfurafine are still unknown. Therefore, we evaluated the partial structures of nalfurafine that are necessary for the selectivity of these two pathways. We assayed the properties of nalfurafine and six nalfurafine analogs (SYKs) using cells stably expressing κORs. The SYKs activated κORs in a concentration-dependent manner with higher EC50 values than nalfurafine. Upon bias factor assessment, only SYK-309 (possessing the 3S-hydroxy group) showed higher selectivity of G-protein-mediated signaling activities than nalfurafine, suggesting the direction of the 3S-hydroxy group may affect the β-arrestin-mediated pathway. In conclusion, nalfurafine analogs having a 3S-hydroxy group, such as SYK-309, could be considered G-protein-biased κOR agonists.


Introduction
Opioid analgesics such as morphine are widely used to improve various forms of pain, including chronic pain, perioperative pain, and cancer pain [1][2][3][4]. ORs belong to the G-protein-coupled receptor (GPCR) family [5] and are classified into three subtypes (µORs, δORs, and κORs), where each type of OR is associated with analgesic effects. Current opioid analgesics mainly bind to µOR to exert their analgesic effects [6]. The typical OR agonists, such as morphine, activate µORs and show strong antinociceptive effects. assays and revealed the structure-activity relationship (SAR) of nalfurafine for κORs [55]. The differences in affinity of each nalfurafine analog for κORs were observed; however, how the structural properties of nalfurafine and its analogs affect the selectivity of the G-protein-and β-arrestin-mediated pathways remains unknown. Accordingly, to mitigate the negative effects of κOR agonists, especially nalfurafine analogs, and effectively utilize them for the treatment of pain, further investigation into the relationship between the structural features of nalfurafine and G-protein-/β-arrestin-mediated signaling activities is required.
Therefore, in the present study, we aimed to evaluate the partial structures of nalfurafine and six nalfurafine analogs (SYK-160, -186, -245, -308, -309, and -406; Figure 1) that are necessary for the selectivity of G-protein-and β-arrestin-mediated pathways. We used the CellKey TM , GloSensor ® cAMP, and PathHunter ® β-arrestin recruitment assays in cells stably expressing κORs. We estimated the G-protein-biased factor of each nalfurafine analog in comparison with nalfurafine to contribute to the development of more useful opioid analgesics. drug because the analgesic and sedative effects were not well separated [47,54]. Even no published clinical evidence indicates the efficacy of nalfurafine for the treatme pain, and nalfurafine is no longer approved as an analgesic. Therefore, in our prev study, we investigated the affinity of nalfurafine and its analogs for κORs using bin assays and revealed the structure-activity relationship (SAR) of nalfurafine for κORs The differences in affinity of each nalfurafine analog for κORs were observed; how how the structural properties of nalfurafine and its analogs affect the selectivity of th protein-and β-arrestin-mediated pathways remains unknown. Accordingly, to mit the negative effects of κOR agonists, especially nalfurafine analogs, and effectively u them for the treatment of pain, further investigation into the relationship betwee structural features of nalfurafine and G-protein-/β-arrestin-mediated signaling acti is required.
Therefore, in the present study, we aimed to evaluate the partial structures o furafine and six nalfurafine analogs (SYK-160, -186, -245, -308, -309, and -406; Figu that are necessary for the selectivity of G-protein-and β-arrestin-mediated pathway used the CellKey TM , GloSensor ® cAMP, and PathHunter ® β-arrestin recruitment assa cells stably expressing κORs. We estimated the G-protein-biased factor of each nalfur analog in comparison with nalfurafine to contribute to the development of more u opioid analgesics.  . Molecular structures of nalfurafine and six nalfurafine analogs. Nalfurafine analogs were divided into two groups according to their structural characteristics. Group A (SYK-160, -186, and -406) includes nalfurafine analogs with a maintained benzene ring. Group B (SYK-245, -308, and -309) includes nalfurafine analogs with a cyclohexene ring converted from the benzene ring.

The Effects of Nalfurafine and Nalfurafine Analogs on the Functions of κORs Using the CellKey TM System
We evaluated the effects of nalfurafine and six nalfurafine analogs (SYK-160, -186, -245, -308, -309, and -406) on κOR activities using the CellKey TM system in HEK293 cells stably expressing Halotag ® -κOR/pGS22F. The CellKey TM system detects the activities of GPCRs, including κORs, as changes in cellular impedance [56]. The E max and EC 50 values were calculated, and we compared them between nalfurafine and its analogs. Nalfurafine and each nalfurafine analog activated κORs in a concentration-dependent manner; however, none of the six analogs exhibited E max (%) values higher than those of nalfurafine ( Figure 2). In contrast, the log EC 50 (M) values of SYK-186 (removed the 3-hydroxy group from nalfurafine in Group A), -245 (removed the 3-hydroxy group and 4,5-ether bridge from nalfurafine, and converted the benzene ring to a cyclohexene ring in Group B), -308 (removed the 4,5-ether bridge from nalfurafine, converted the benzene ring to a cyclohexene ring, and added a 3R-hydroxy group in Group B), and -406 (removed the 3-hydroxy group and 4,5-ether bridge from nalfurafine in Group A) were significantly increased compared to those of nalfurafine (Table 1).

The Effects of Nalfurafine and Nalfurafine Analogs on the Functions of κORs Using the CellKey TM System
We evaluated the effects of nalfurafine and six nalfurafine analogs (SYK-160, -186, -245, -308, -309, and -406) on κOR activities using the CellKey TM system in HEK293 cells stably expressing Halotag ®️ -κOR/pGS22F. The CellKey TM system detects the activities of GPCRs, including κORs, as changes in cellular impedance [56]. The Emax and EC50 values were calculated, and we compared them between nalfurafine and its analogs. Nalfurafine and each nalfurafine analog activated κORs in a concentration-dependent manner; however, none of the six analogs exhibited Emax (%) values higher than those of nalfurafine ( Figure 2). In contrast, the log EC50 (M) values of SYK-186 (removed the 3-hydroxy group from nalfurafine in Group A), -245 (removed the 3-hydroxy group and 4,5-ether bridge from nalfurafine, and converted the benzene ring to a cyclohexene ring in Group B), -308 (removed the 4,5-ether bridge from nalfurafine, converted the benzene ring to a cyclohexene ring, and added a 3R-hydroxy group in Group B), and -406 (removed the 3-hydroxy group and 4,5-ether bridge from nalfurafine in Group A) were significantly increased compared to those of nalfurafine (Table 1).

Figure 2.
Effect of nalfurafine and six nalfurafine analogs on κORs observed using the CellKey TM system. The cells expressing κORs were treated with nalfurafine (positive control) and six nalfurafine analogs (Group A: SYK-160, -186, -406; Group B: SYK-245, -308, -309) at concentrations of 10 −13 -10 −5 M, and changes in impedance (ΔZiec) were measured using the CellKey TM system. Concentration-response curves were prepared by calculating ΔZiec relative to the data obtained for the positive control: 10 −7 M nalfurafine. All data points are presented as means ± standard error of the mean (SEM) (n = 3-6). Nalfurafine was used as the positive control. Emax (%) and log EC50 (M) values (means ± SEM) were calculated according to the results shown in Figures 2-4. Statistical comparisons were made using GraphPad Prism 9 software and are expressed as means ± SEM. Differences between the means were analyzed with one-way analysis of variance (ANOVA) or t-tests. One-way ANOVA was followed by Bonferroni post hoc analysis. Significant levels are ** p < 0.01 and *** p < 0.001 compared with nalfurafine. The number of samples of EC50 and Emax are indicated as follows: n = 3-6  and are expressed as means ± SEM. Differences between the means were analyzed with one-way analysis of variance (ANOVA) or t-tests. One-way ANOVA was followed by Bonferroni post hoc analysis. Significant levels are ** p < 0.01 and *** p < 0.001 compared with nalfurafine. The number of samples of EC 50 and E max are indicated as follows: n = 3-6 (the CellKey TM assay), n = 3-9 (the GloSensor ® cAMP assay), and n = 5-8 (the PathHunter ® recruitment assay).

The Effects of Nalfurafine Analogs on the Intracellular cAMP Levels Evaluated Using the GloSensor ® cAMP Assay
We evaluated the actions of test compounds on κOR-induced G-protein signaling by measuring the intracellular cAMP levels using HEK293 cells stably expressing Halotag ® -κOR/pGS22F. The E max and EC 50 values were calculated using nalfurafine as a positive control, and the six nalfurafine analogs caused a concentration-dependent decrease in cAMP levels ( Figure 3). In detail, there were no nalfurafine analogs that showed E max (%) values higher than nalfurafine, and the log EC 50 (M) values of all nalfurafine analogs were significantly increased compared to those of nalfurafine (Table 1). These results suggested that the six nalfurafine analogs used in this study showed lower G-protein-mediated signaling activities than nalfurafine. ment assay).

The Effects of Nalfurafine Analogs on the Intracellular cAMP Levels Evaluated Using the GloSensor ®️ cAMP Assay
We evaluated the actions of test compounds on κOR-induced G-protein signaling by measuring the intracellular cAMP levels using HEK293 cells stably expressing Halotag ®️ -κOR/pGS22F. The Emax and EC50 values were calculated using nalfurafine as a positive control, and the six nalfurafine analogs caused a concentration-dependent decrease in cAMP levels ( Figure 3). In detail, there were no nalfurafine analogs that showed Emax (%) values higher than nalfurafine, and the log EC50 (M) values of all nalfurafine analogs were significantly increased compared to those of nalfurafine (Table 1). These results suggested that the six nalfurafine analogs used in this study showed lower G-protein-mediated signaling activities than nalfurafine.

Effects of Nalfurafine Analogs on β-Arrestin Recruitment Using the PathHunter ® Recruitment Assay
To evaluate the actions of nalfurafine and six nalfurafine analogs on κOR-induced βarrestin signaling, the PathHunter ® β-arrestin recruitment assay was performed using U2OS cells stably expressing κORs (DiscoverX, Fremont, CA, USA). Nalfurafine and each nalfurafine analog induced β-arrestin recruitment to κORs in a concentration-dependent manner ( Figure 4). We calculated the Emax and EC50 values of these compounds, and no nalfurafine analogs showed Emax (%) values higher than nalfurafine. However, the log EC50 (M) values of all nalfurafine analogs were increased, except SYK-160, which has a nonsignificant increase compared to those of nalfurafine.

The Selectivity of G-Protein-and β-Arrestin-Mediated Pathways (G-Protein-Biased Factors)
The selectivity of G-protein-and β-arrestin-mediated pathways is indicated as a Gprotein-biased factor representing the ratio of the value for G-protein signaling divided by that of β-arrestin signaling [57]. A biased factor > 1 indicates a preference for G-proteinmediated signaling activities, whereas a biased factor < 1 indicates a preference for the recruitment of β-arrestin as β-arrestin-mediated signaling activities compared to the control compound [56]. Subsequently, we estimated the G-protein-biased factor of each nalfurafine analog compared to nalfurafine to identify G-protein-biased analogs. The present study calculated G-protein-biased factors using data from the GloSensor ®️ cAMP assay (Gprotein-mediated signaling) and PathHunter ® recruitment assay (β-arrestin-mediated signaling). As shown in Table 2, the G-protein-biased ratio of SYK-309 was significantly higher than that of nalfurafine (mean ± SEM: 4.46 ± 1.87, p = 0.0055, Table 2). These results indicated that, compared to nalfurafine, SYK-309 was the only G-protein-biased κOR agonist among the six nalfurafine analogs. Table 2. G-protein-biased factors of nalfurafine analogs for G-protein and β-arrestin coupling.

Effects of Nalfurafine Analogs on β-Arrestin Recruitment Using the PathHunter ® Recruitment Assay
To evaluate the actions of nalfurafine and six nalfurafine analogs on κOR-induced β-arrestin signaling, the PathHunter ® β-arrestin recruitment assay was performed using U2OS cells stably expressing κORs (DiscoverX, Fremont, CA, USA). Nalfurafine and each nalfurafine analog induced β-arrestin recruitment to κORs in a concentration-dependent manner (Figure 4). We calculated the E max and EC 50 values of these compounds, and no nalfurafine analogs showed E max (%) values higher than nalfurafine. However, the log EC 50 (M) values of all nalfurafine analogs were increased, except SYK-160, which has a non-significant increase compared to those of nalfurafine.

The Selectivity of G-Protein-and β-Arrestin-Mediated Pathways (G-Protein-Biased Factors)
The selectivity of G-protein-and β-arrestin-mediated pathways is indicated as a Gprotein-biased factor representing the ratio of the value for G-protein signaling divided by that of β-arrestin signaling [57]. A biased factor > 1 indicates a preference for G-proteinmediated signaling activities, whereas a biased factor < 1 indicates a preference for the recruitment of β-arrestin as β-arrestin-mediated signaling activities compared to the control compound [56]. Subsequently, we estimated the G-protein-biased factor of each nalfurafine analog compared to nalfurafine to identify G-protein-biased analogs. The present study calculated G-protein-biased factors using data from the GloSensor ® cAMP assay (G-proteinmediated signaling) and PathHunter ® recruitment assay (β-arrestin-mediated signaling). As shown in Table 2, the G-protein-biased ratio of SYK-309 was significantly higher than that of nalfurafine (mean ± SEM: 4.46 ± 1.87, p = 0.0055, Table 2). These results indicated that, compared to nalfurafine, SYK-309 was the only G-protein-biased κOR agonist among the six nalfurafine analogs. Table 2. G-protein-biased factors of nalfurafine analogs for G-protein and β-arrestin coupling. The parameters were calculated from the same agonist concentration-response curves used to estimate EC 50 and E max values in Figures 2 and 3, and in Table 1, using the method described by Ehlert and colleagues [58][59][60]. The prototype of the selective κOR agonist, nalfurafine, was designated as a standard reference ligand. The bias factor of G-protein signaling for a given ligand is defined as the ratio of the intrinsic activity (RA i-G ) divided by RA i-b . The G-protein-biased ratios (means ± SEM) were calculated according to the results shown in Table 1, ** p < 0.01, compared to bias factor of 1 by t-test.

Discussion
Here, we evaluated the effects of nalfurafine and six nalfurafine analogs (SYK-160, -186, -245, -308, -309, and -406) on κOR-activated intracellular signaling using the CellKey TM , GloSensor ® cAMP, and PathHunter ® β-arrestin recruitment assays. Our results revealed that all tested compounds activated κOR-mediated intracellular signaling in a concentrationdependent manner as full κOR agonists. Furthermore, most of the EC 50 values of these test compounds were higher than nalfurafine. In addition, similar results were obtained in our CellKey TM assay and GloSensor ® cAMP assay and their correlation seemed to be high. These results suggest that in an impedance assay using the CellKey TM system, the results of impedance changes were reflected mostly with changes in cAMP levels, but not changes in β-arrestin activity. Our previous studies also showed a similar pattern to the present results [61,62]. Our previous study examined the binding affinity of six nalfurafine analogs (the same ones used in the present study) for κORs; nitrogen, with an Ncyclopropylmethyl substituent, and 6-amide side chains were indispensable for nalfurafine to bind to κORs, and the phenol ring (3-hydroxy group) was also important for increasing the κOR binding affinity. Compared to our present study, the binding ability K i (nM) value of each nalfurafine analog for κORs tended to correlate with the EC 50 values of CellKey TM , cAMP, and β-arrestin recruitment assays [55]. Moreover, the results of one analog, SYK-309, indicated that there was a significant difference in the ratio of G-protein-mediated signaling to β-arrestin-mediated signaling in our present study.
The κOR agonists have been proposed as antinociceptive drugs in humans [24,46,63,64]. These agonists can independently activate multiple signaling mechanisms, making it diffi-cult to screen them using one assay [65]. For this reason, we used three assays in our present study. Schattauer S.S. et al. showed that nalfurafine was a G-protein-biased κOR agonist compared to other κOR agonists such as (-)-U50488H [41]. Nalfurafine exerted anti-scratch and analgesic effects without adverse events such as sedation, motor incoordination, or conditioned place aversion in mice [66]. Many investigations indicated that antinociception induced by κOR agonists is caused by the G-protein-mediated pathway [67], whereas unfavorable effects are caused by the β-arrestin-mediated pathway [68][69][70], suggesting that G-protein-biased κOR agonists could lead to the development of safer and more effective opioid analgesics. Indeed, nalfurafine was initially developed as an analgesic; however, the analgesic and sedative effects were not well separated at analgesic doses [47,54]. Taking the previous studies into consideration, we focused on nalfurafine and investigated the relationships between κOR signaling selectivity and the structural features of nalfurafine using six nalfurafine-based analogs.
Among the nalfurafine analogs in Group A (maintained benzene ring), none had E max values that significantly exceeded those of nalfurafine in both G-protein-and β-arrestinmediated signaling. In contrast, for G-protein-mediated signaling, SYK-160 (removed the 4,5-ether bridge), SYK-186 (removed the 3-hydroxy group), and SYK-406 (removed both the 4,5-ether bridge and 3-hydroxy group) caused significant increases in EC 50 values. However, for β-arrestin-mediated signaling, the EC 50 value of SYK-160 was not significantly changed, whereas those of SYK-186 and SYK-406 were significantly increased. Compared to nalfurafine, the G-protein-biased ratio of SYK-160 was decreased. Still, there was no significant change in the ratios of the Group A compounds (Table 2). Therefore, these data suggest that the 4,5-ether bridge and 3-hydroxy group on the benzene ring were important to activate both G-protein-and β-arrestin-mediated signaling.
Moreover, our present study showed that SYK-309, but not SYK-308, significantly increased the G-protein-biased ratio compared to nalfurafine ( Table 2). As a cyclohexene ring is not an aromatic ring, the carbon bound to the 3-hydroxy group is a stereogenic center. Therefore, SYK-308 and SYK-309 possess a 3R and 3S configuration, respectively. Between SYK-308 and -309, the direction of the 3-hydroxy group was the only structural difference; however, there were significant differences in the EC 50 values of β-arrestinmediated signaling (PathHunter ® recruitment assay, p = 0.031), but not in G-proteinmediated signaling (GloSensor ® cAMP assay, p = 0.576). Therefore, our present study, for the first time, suggests that the difference in the direction of a hydroxy group at the 3position of the cyclohexene ring may cause a change in the selectivity of β-arrestin-mediated signaling, and the direction of the 3S-hydroxy group could be one of the important key factors for G-protein-biased κOR signaling.
Molecular docking studies of nalfurafine and SYK-186 (removal of the 3-hydroxy group) into κORs have shown that the 3-hydroxy group of the phenolic moiety of nalfurafine interacted with residues Y3.33, K5.39, and H6.52 in κORs via water-mediated hydrogen bonds. In contrast, SYK-186 did not interact with these residues in κORs [57].
However, it is unknown which of the residues in κORs interacts with SYK-309. Therefore, considering the structural features of nalfurafine and the expectation for the development of novel κOR agonists, further investigation with a 3-D docking model using SYK-309 could elucidate the direction of the hydroxy group at the 3-position of nalfurafine analogs that induces G-protein-biased signaling.
There are some limitations in the present study. Since this was an in vitro study, we cannot clinically evaluate the analgesic effects or side effects of the nalfurafine analogs compared to nalfurafine. As a result, we cannot conclude whether SYK-309 can maintain analgesic effects and reduce any unfavorable side effects mediated by the β-arrestinmediated pathway. In addition, whether G-protein-biased κOR agonists have safer and more beneficial profiles than opioid analgesics is still under discussion. Therefore, further studies in different models are necessary to develop more efficacious opioids without any negative impacts on disorders such as chronic pain and pruritus.
Several studies, particularly by Laura Bohn, claim that arrestin is primarily responsible for the adverse effects of ORs; however, numerous recent studies contradict this assertion [34,71]. Therefore, it is necessary to analyze not only analgesic effects, but also sedative effects of SYK-309, as well as its property as a G-protein-biased agonist against κORs. Moreover, we measured the κOR activities (the G-protein-mediated and β-arrestin-mediated signaling) for only the κOR agonists, the nalfurafine and nalfurafine analogs, in this study. Therefore, it is also necessary that the κOR activities in other morphinan or benzomorphan derivatives which have the A ring modified like the SYK-309 are measured and have their profiles compared with nalfurafine and nalfurafine analogs in the future.
Opioid analgesics, especially µOR agonists, are used to treat pain; however, their usage is sometimes complicated by detrimental side effects [7]. Therefore, developing novel opioids with fewer adverse events is strongly desirable. Recent research indicated that functionally selective κOR agonists elicited neither addictive nor adverse effects [72], and, subsequently, several groups have screened for G-protein-biased κOR agonists [73][74][75][76]. This study showed that the direction of the 3-hydroxy group in nalfurafine is crucial in inducing G-protein-biased signaling. However, to clinically introduce κOR agonists as painkillers, further investigation of the structural properties of nalfurafine-which selectively activate G-protein-mediated pathways via κORs-is necessary.

Cell Lines
We amplified Halotag ® -fused κORs (Halotag ® -κOR, from Kazusa DNA Research Institute, Chiba, Japan) with the pGlosensor TM -22F plasmid (pGS22F) from Promega (Madison, WI, USA), following the manufacturer's instructions. Human embryonic kidney 293 (HEK293) cells were obtained from the American Type Culture Collection (ATCC ® , Manassas, VA, USA), and stably expressing Halotag ® -κORs were generated by transfection of the constructed plasmids using the Lipofectamine reagent (Life Technologies, Carlsbad, CA, USA), which were selected based on OR activity measured by the CellKey TM assay or the cAMP assay with Glosensor ® .

Functional Analysis of ORs Using the CellKey TM System
We examined the effects of nalfurafine and nalfurafine analogs on κORs by the CellKey TM assay system, as described previously [56,79]. We seeded cells at a density The change in impedance of an induced extracellular current (dZiec) in each well was measured for 25 min, following a 5 min baseline measurement. The magnitude of change in the dZiec value was defined as ∆Ziec. The value for nalfurafine analogs was calculated as a percentage using the highest value for nalfurafine (positive control).

Intracellular cAMP Levels Measured with the GloSensor ® cAMP Assay
We performed the GloSensor ® cAMP assay as described previously [30,79,80]. In brief, cAMP accumulation was analyzed using cells stably expressing Halotag ® -KOR/pGS22F. We seeded the cells at 4.0 × 10 4 cells/well in 96-well clear-bottom white plates (Corning, Corning, NY, USA) and then incubated them for 24 h. After washing the cells with the CellKey TM buffer without BSA, the cells were equilibrated with the diluted GloSensor ® reagent at room temperature for 2 h, and the baseline fluorescence intensity was measured for 15 min. After the baseline measurement, cells were treated with the test compounds for 10 min, after which forskolin (3.0 × 10 −6 M) was added. The fluorescence intensity was measured every 2.5 min for 30 min using Synergy TM H1 (Bio Tek Instruments Inc., Winooski, VT, USA); time-fluorescence curves and the area under the curve (AUC) values of time-fluorescence intensities were calculated. The responses of test compounds were expressed as the AUC of each test compound subtracted from that of the negative control sample (forskolin alone). Data were transformed from each well as the percentage (%) of intracellular cAMP inhibition and calculated by dividing the corrected AUC by those of the standard sample. The standard sample was nalfurafine (10 −7 M) for Halotag ® -KOR/pGS22F.

β-Arrestin Recruitment Assay with PathHunter ®
This was performed as described previously [81]. In brief, U2OS OPRM1, CHO-K1 OPRD1, or U2OS OPRK1 cells were seeded at a density of 1.0 × 10 4 cells/well in 96-well clear-bottom white plates and incubated for 48 h. The cells were stimulated for 180 min at 37 • C under 5% CO 2 , and the PathHunter ® working detection solution was added. The luminescence intensity was measured using the FlexStation 3 (Bio Tek Instruments Inc., Winooski, VT, USA) for 1 h at 25 ± 3 • C. Data are expressed as the maximum signal intensity of each test compound as a percentage of the maximum signal intensity of the positive control.

The Estimated Intrinsic Reactive Activity (RA i ) and Biased Factors
According to the method developed and refined by Ehlert and colleagues [58][59][60], the G-protein-biased factor was estimated. In brief, each agonist's intrinsic reactive activity (RA i ) was estimated by global nonlinear regression analysis [59,82,83]. The RA i of each nalfurafine analog was estimated from the concentration-response curves used to estimate EC 50 and E max values ( Table 1). The G-protein-biased factor was defined as the ratio of the RA i g value divided by RA i-B (Table 2).

Statistical Analysis and Approval for the Study
Data analyses and concentration-response curve fitting were performed using Graph-Pad Prism 9 (GraphPad Software, San Diego, CA, USA). Data are presented as means with the standard error of the mean (SEM) for at least three independent experiments. Statistical analysis was performed using a one-way ANOVA, followed by the Bonferroni multiple comparison tests or t-tests. A value of p < 0.05 was considered statistically significant. All analyses and experiments were approved and performed in accordance with the Guide for Genetic Modification Safety Committee, National Cancer Center, Japan.

Conclusions
The present study revealed that the direction of the 3-hydroxy group of nalfurafine may be the partial structure inducing G-protein-biased signaling via the weakening of βarrestin-mediated signaling. Therefore, nalfurafine analogs having a 3"S"-hydroxy group, such as SYK-309, could be considered κOR agonists with a pharmacological profile that selectively activates the G-protein-mediated pathway. This evaluation of the structureactivity relationship is expected to help the development of novel selective κOR agonists.